Application of the PSO for the construction of a 3-axis stable magnetically actuated satellite angular motion

Abstract

Particle swarm optimization (PSO) is a non-gradient biologically inspired global optimization method. It utilizes direct calculation of the optimized cost function during the numerical simulation of the process in question. Significantly complex and implicit functions that prohibit clear gradient expression are effectively handled by PSO. This feature is utilized in the paper to construct specific angular trajectory of a magnetically actuated satellite. A cost function is suggested that confines the control torque corresponding to this trajectory to the plane orthogonal to the geomagnetic induction vector. Optimal trajectory coefficients are found by PSO. The control structure is based on the Lyapunov feedback approach and explicitly accommodates the gravitational and aerodynamic torques influence. PSO procedure provides the optimal control gains on this stage. Stability of the resulting motion is demonstrated. Numerical simulation example with different disturbance factors illustrates the expected stabilization accuracy of the order of few degrees.